EP1178301B1

EP1178301B1 - Defect marking method and device

Info

Publication number: EP1178301B1
Application number: EP00909637A
Authority: EP
Inventors: Mitsuaki NKK Corporation UESUGI; Shoji NKK Corporation YOSHIKAWA; Masaichi NKK Corporation INOMATA; Tsutomu NKK Corporation KAWAMURA; Takahiko NKK Corporation OSHIGE; Hiroyuki NKK Corporation SUGIURA; Akira NKK Corporation KAZAMA; Tsuneo NKK Corporation SUYAMA; Yasuo NKK Corporation KUSHIDA; Shuichi Fukuyama Kyodokiko Corporation HARADA; Hajime NKK Corporation TANAKA; Osamu NKK Corporation UEHARA; Shuji NKK Corporation KANETO; Masahiro NKK Corporation IWABUCHI; Kozo NKK Corporation HARADA; Shinichi NKK Corporation TOMONAGA; Shigemi NKK Corporation FUKUDA
Original assignee: JFE Steel Corp
Current assignee: JFE Steel Corp
Priority date: 1999-03-18
Filing date: 2000-03-15
Publication date: 2007-10-31
Anticipated expiration: 2020-03-15
Also published as: EP1178301A1; US7599052B2; EP1857811A2; CA2676748C; WO2000055605A1; KR100458048B1; CA2365879C; US20020154308A1; CA2365879A1; DE60036939T2; KR20020011372A; CA2676748A1; KR20040028944A; DE60036939D1; US20070052964A1; US20090086209A1; EP1857811A3; US7248366B2; US7423744B2; KR100568973B1

Description

FIELD OF THE INVENTION

The present invention relates to a method for marking defect on a steel sheet in a steel making process line to a flaw inspection device, and to a defect marking device (see JP-A-9-166549 ).

DESCRIPTION OF RELATED ARTS

Cold-rolled steel sheets manufactured by cold-rolling are subjected to inspection of surface defects over the whole length of coil thereof for quality assurance. JP-A-5-196581 , (the term "JP-A" referred herein signifies the "Unexamined Japanese patent publication"), discloses a method for detecting surface defects and internal defects of steel sheets. According to the disclosure, detection of the surface defects on a steel sheet is conducted by scanning the surface of steel sheet which is running through a manufacturing line, in width direction thereof by laser light, by converting the reflected light to voltage intensity using a photoelectric transfer device such as CCD element, then by judging the presence/absence and the degree of the defect based on the voltage signals. The internal defects of a steel sheet are detected by computing the defect depth in the thickness direction of the steel sheet and the defect size using a magnetic particle tester. Usually, the result of the defect inspection is displayed on a CRT or the like as information in terms of defect position, defect name, defect grade, and the like, or is printed in a document.
It is impossible to obtain products completely free from defects. Consequently, the products are shipped after removing portions of harmful defect for the buyer concerned on the basis of the defect information displayed on a CRT or the like in the manufacturing line, or after removing the portions of harmful defect for the buyer concerned by applying re-inspection in succeeding stage on the basis of the above-described defect information. Alternatively, a document of the above-described harmful detect information is submitted to the buyer concerned, together with the coil that contains harmful defect portions, thus letting the buyer remove the harmful defects.
In the case that the harmful defect portions are removed in the manufacturing line or in succeeding stage, since there is no definite standard of the degree of harm for the surface defects, the removal of harmful defect portions is practiced in an excessive action from the standpoint of quality assurance. Also there are cases of not-removing harmful defects caused from a miss-judgment such as overlooking and from a state of very close to injudgicable defect. Furthermore, removal of harmful defect portions raises problems such as reduction in the coil weight and reduction in the work efficiency of the buyer.
On the other hand, the buyer needs to work on coils while referring the documented data of defect information, which induces troublesome work and, in some cases, may result in treatment of coils leaving defects non-detected.
JP-A-4-291138 discloses a marking device that sprays a paint on flawed portions of steel sheets. According to the disclosure, marking is done by spraying a paint to flawed portions of the steel sheets detected by a flaw detection device, thus enabling the buyer to readily identify the flawed portion on re-inspection by the buyer.
Since, however, the method for marking according to JP-A-4-291138 does not give confirmation whether the marking was correctly given or not, and an abnormal marking induces further troubles to the buyer. In addition, paint spray generates shade of color on marking, which may induce dents, in an area with large quantity of applied paint, even at normal portions after the paint is dried. For the case of spray marking, an oiled steel sheet cannot leave any marking on the surface thereof because the paint is sprayed on an oil film, though that kind of problem does not occur on a steel sheet free of applied oil. When marking is given on all of the flawed portions, the marking is also given to the defects of harmless to the buyer, which induces disadvantages including reduction in work efficiency.
There are inspection methods for surface defects, disclosed in JP-A-58-204353 , JP-A-60-228943 , JP-A-8-178867 , JP-A-57-166532 , and JP-A-9-166552 . All of these disclosed methods aim to detect flaws having significant surface irregularity or to detect flaws with the presence of foreign matter such as oxide film. Thus, for pattern-like scabbed flaws or the like which do not have significant surface irregularity, these methods cannot surely identify all the flaws.
As means for applying marking to flawed portions and singular parts generated on a metal material, there are commercially available apparatuses such as ink jet printer and ink dot marking device.
When an ink jet printer is used, kinds and colors of ink are limited because special inks are required owing to various conditions such as charging the inks. Nevertheless, when the manufactured metal materials are used for automobile steel sheets, ink performance and color may be specified for convenience of customer's inspection.
For example, if an ink is under limitations such that the ink should have excellent quick drying property, the ink should not be blotting on applying oil, and the ink should be blue, respective special inks should be developed to satisfy the customer's conditions. The development of that special inks needs long time and much money, which is difficult to practically respond to the request.
In addition, since the ejection part of the ink is necessary to maintain clean, significant cost and time should be consumed for maintenance. Accordingly, when an ink jet printer is used, a special ink has to be used, so that color and kind of the ink cannot readily be changed.

SUMMARY OF THE INVENTION

It is an object of the present invention to provide a method for defect marking that readily and surely identifies harmful defects, and an apparatus therefor.
In one aspect, the invention provides a flaw inspection device according to claim 1.
In another aspect, the invention provides a defect marking device according to claim 2.
In a further aspect, the invention provides a method according to claim 3. Optional features of the method are defined by claim 4.

BRIEF DESCRIPTION OF THE DRAWINGS

Fig. 1 is a block diagram of an example of the devices relating to the present invention.
Fig. 2 is a plan view of an example of metal strip relating to the present invention.
Fig. 3 is a schematic drawing of an example of rough structure of a surface flaw inspection device for the devices relating to the present invention.
Fig. 4 is a cross sectional schematic drawing of a surface flaw detection device relating to the present invention.
Fig. 5 illustrates an arrangement of camera units along the metal width direction, equipped in the surface flaw inspection device, relating to the present invention.
Fig. 6 illustrates an arrangement of cameras equipped in a single camera unit, relating to the present invention.
Fig. 7 is a block diagram of another example of the devices relating to the present invention.
Fig. 8 is a block diagram of an example of the signal processing section of the devices relating to the present invention.
Fig. 9 is a block diagram of further example of the devices relating to the present invention.
Fig. 10 (a) through (c) illustrates examples of light intensity signals observed by a device relating to the present invention.
Fig. 11 (a) through (c) illustrates other examples of light intensity signals observed by a device relating to the present invention.
Fig. 12 (a) through (d) illustrates the method for manufacturing alloyed zinc plated steel sheet and shows detailed cross sectional views of the sheet, relating to the present invention.
Fig. 13 shows a schematic cross sectional view of the tempered part and the non-tempered part on the surface of metal strip after temper-rolling, illustrating the relation between the incident light and the reflection light, relating to the present invention.
Fig. 14 (a) through (c) shows angle distribution of reflected light at the tempered part and the non-tempered part, relating to the present invention.
Fig. 15 shows cross sectional views of an alloy zinc plated steel sheet to illustrate the progress of occurrence of scab, relating to the present invention.
Fig. 16 (a) through (c) shows angle distribution of specular reflection component and specular-diffuse reflection component, at scabbed portion and mother material, relating to the present invention.
Fig. 17 (a) through (c) shows the relation between the normal angle to micro-area element and the area percentage, at scabbed portion and mother material on the inspection plane, relating to the present invention.
Fig. 18 shows the relation between angles of the incident light, the reflection light, and the like, on a micro-area element on the inspection plane, relating to the present invention.
Fig. 19 (a) and (b) show the relation between the normal angle to a micro-area element and the weight function, relating to the present invention.
Fig. 20 (a) and (b) shows the relation between individual incident lights emitted from various positions on the linear diffusion light source and the responding incident positions on the inspection plane, relating to the present invention.
Fig. 21 (a) and (b) shows the polarized state of reflection light coming from a micro-area element in the case that each incident light coming from the linear diffusion light source is polarized, relating to the present invention.
Fig. 22 illustrates the reflection light coming from a micro-area element in the case that the incident light coming from center part of the linear diffusion light source is polarized, relating to the present invention.
Fig. 23 illustrates the reflection light coming from a micro-area element in the case that the incident light coming from a part other than the center part of the linear diffusion light source is polarized, relating to the present invention.
Fig. 24 illustrates the relation between the normal angle to micro-area element and the elliptic polarized light of the reflected light, relating to the present invention.
Fig. 25 shows the relation between the normal angle to micro-area element and the weight function, relating to the present invention.
Fig. 26 shows the relation between the normal angle to micro-area element and the weight function at various analyzing angles, relating to the present invention.
Fig. 27 shows the relation between the normal angle to micro-area element on the inspection plane and the area percentage, relating to the present invention.

The first aspect of the invention that will be described is a defect marking device according to claim 2.
This device receives light reflected from the surface of the metal strip by two or more of light-receiving part, having different optical conditions such as polarization condition to each other, and analyzes the optical properties from the received result. Then, the signal processing section of the flaw inspection means gives judgment on normal part and abnormal part, or judgment on the surface flaw, on the surface of the metal strip based on thus obtained optical properties. For the part judged as the surface flaw, marking is applied using a specified method such as printing, carved stamping, and drilling. The position for marking can be selected by tracking the position of or nearby the surface flaw using a tracking means or the like.
The following is the description about the mode of optical reflection on the surface of steel sheet, which is a target of the inspection by the surface flaw inspection device according to the present invention, relating to microscopic surface irregularity on the surface of the steel sheet. Generally, the microscopic irregularity on the surface of steel sheet, which is originally significantly rough, improves its flatness by strong rolling by a roll during temper rolling (tempering), while other portions leave their irregular shape because the roll of the temper rolling does not contact thereto.
For example, in the case of alloyed galvanized steel sheet, the cold-rolled steel sheet 101, the mother material, is subjected to hot dip galvanizing as shown in Fig. 12(a), then passes through an alloying furnace. During the passage, the iron element of the mother material steel sheet diffuses into the zinc of the plating layer to generally form columnar alloy crystals 103 as shown in Fig. 12(c). When the steel sheet is subjected to temper rolling as shown in Fig. 12(b), the particularly projected portions of the columnar crystals 103 are collapsed in flat shape, (tempered part 106), as shown in Fig. 12(d), while leaving other portions (non-tempered part 107) as in the columnar crystal shape.
Fig. 13 is a model illustrating what kinds of optical reflections occur on that type of steel sheet surface. The incident light 108 coming into the portion collapsed by temper rolling, (tempered part 106), gives specular reflection to the direction of regular reflection to the steel sheet. On the other hand, the reflection direction of the incident light coming into the portion which leaves original columnar crystals not collapsed by the temper rolling, (non-tempered part 107), does not necessarily coincide with the regular reflection direction to the steel sheet, though it is reflected in specular manner on individual micro-area elements on the columnar crystal surface in microscopic view.
Therefore, the distribution of reflection light angles at tempered part and non-tempered part becomes to Fig. 14(a) and Fig. 14(b) in macroscopic view, respectively. That is, (a) at the tempered part 106, a specular reflection 109 having a sharp distribution in the regular reflection direction to the steel sheet occurs, and (b) at the non-tempered part, a reflection 110 having a broad range responding to the angle distribution on micro-area elements on the surface of columnar crystals appears. Hereinafter the former is referred to as the specular reflection, and the latter is referred to as the specular-diffuse reflection. The actually observed distribution of reflection angles is the sum of the angle distribution of specular reflection and the angle distribution of specular-diffuee reflection responding to each area percentage of the tempered part and the non-tempered part, as shown in Fig. 14(c).
The above-given description deals with an alloyed galvanized steel sheet. However, the description is generally applicable to other steel sheets generating flat portions by temper rolling.
The following is the description about the optical reflection characteristics of flaw called the pattern-like scab; and which has no significant surface irregularity, which is a target of the present invention. For example, as seen in Fig. 15, a scab 111 appeared on an alloyed hot dip galvanized steel sheet 104 exists in an original plate 101 of cold-rolled steel sheet before plating, on which a plating layer 102 is applied, and further the alloying proceeds by diffusion of the iron in the mother material.
Compared with mother material, the scabbed portion generally differs in plating thickness and in degree of alloying. As a result, for example, in the case that the plating layer thickness at the scabbed portion becomes thick and that the scab is convex against the mother material, the temper rolling increases the area of tempered part than that of non-tempered part. Inversely, if the scabbed portion is concave against the mother material, the scabbed portion does not touch the temper rolling roll, and the non-tempered part occupies large portions. If the alloying at scabbed portion is shallow, the angle distribution of micro-area elements is enhanced in the normal direction to the steel sheet, and the diffusion performance becomes weak.
The following is the description about the appearance of pattern-like scabs depending on the difference in surface property of the scabbed portion and of the mother material. When the difference between the scabbed portion and the mother material is classified depending on the above-described modified model of plating surface during temper rolling, three kinds of groups appear as shown in Fig. 17.

(a) In a scabbed portion (solid line), the area percentage of the tempered part and the angle distribution on micro-area elements in the non-tempered part differ from those in the mother material (broken line). The tempered part corresponds to the normal angle ξ = 0, indicating the peak in the figure. The peak height (area percentage) differs in the scabbed portion and the mother material. The non-tempered part corresponds to the other parts (slope), and, in the figure, the distribution of the area percentage differs in the scabbed portion and the mother material. The slope part reflects the angle distribution on micro-area elements in the non-tempered part.
(b) Although the area percentage of the tempered part differs between the scabbed portion and the mother material, the angle distribution on micro-area elements in the non-tempered elements does not differ from each other. The figure shows different peak height in scabbed portion and in mother material. However, the slope shape agrees to each other.
(c) Although the angle distribution on micro-area elements in the non-tempered part differs between the scabbed portion and the mother material, the area percentage in the tempered part does not differ to each other. The figure shows different peak height in scabbed portion and in mother material. However, the slope shape differs from each other.

That difference in the area percentage of the tempered part and in the angle distribution micro-area elements is observed as the difference in the angle distribution of reflected light quantity, as shown in Fig. 16.
If the area percentage of the tempered part shows a difference, (as in the case of above-described (a) and (b)), the angle distribution of the reflected light quantity becomes that on the scabbed portion 111a and on the mother material 112a, as shown in Figs. 16 (a) and (b). The difference is observed in the direction that the angle distribution becomes a peak, or the direction of regular reflection. If the area percentage of the tempered part in the scabbed portion is larger than that in the mother material, (Figs. 16(a) and (b), and Figs. 17(a) and (b)), the scab is seen bright from the regular reflection direction. And, if the tempered percentage in the scabbed portion is less than that in the mother material, the scab is seen dark from the regular reflection direction.
If there is no difference in the area percentage of the tempered part, (in the case of above-described (c)), the observation from the normal reflection direction to the steel sheet cannot see the scab. Nevertheless, if there is a difference in the diffusion property of the components of specular-diffuse reflection, the flaw can be seen from the diffusional direction at an off-peak angle distribution, as shown in Fig. 16(c). For example, when the diffusional property of the components of specular-diffuse reflection is small, generally the scab is viewed bright from a diffusional direction relatively near to the regular reflection, and the brightness gradually becomes weak with off-setting from the regular reflection direction, and finally, the difference between the scabbed portion and the mother material becomes none at a certain angle, thus the observation at around this angle is no more possible. Further off-setting from the regular reflection angle allows the observation of scab in dark color.
To identify and detect that pattern-like scab from the mother material, it is necessary to investigate the angle of micro-area elements for identifying the reflection light. For example, as in the case of Figs. 16(a) and (b), the detection of difference between the scabbed portion and the mother material in the regular reflection direction means the determination of the ξ= 0 angle distribution among the angle distribution in micro-area elements, shown in Fig. 17, thus detecting the difference between the scabbed portion and the mother material.
When the identification at ξ= 0 angle distribution is described in terms of arithmetic expression, a function S(ξ) shown in Fig. 17 is multiplied with a function that signifies an identification characteristic expressed by a delta function δ(ξ) shown in Fig. 19 (a), (hereinafter referred to simply as "weight function"), then the product is integrated. Furthermore, for example, at an incident light angle of 60 degrees, the observation at 40 degrees, or offsetting by 20 degrees, means that the reflection on a plane (micro-area element) offsets by 10 degrees of normal angled ξ. This corresponds to the use of a weight function of δ(ξ + 10), as seen in Fig. 19 (b). The relation between the reflection angle and the normal angle ξ to a micro-area element is calculated from Fig. 18.
According to the consideration, the identification of reflected light from an angle of micro-area element corresponds to the design of a weight function. The weight function is not necessarily a delta function, and it may have a certain width.
Based on the concept, when the scabs having respective area percentages expressed by Figs. 17 (a), (b), and (c) are identified separately from the mother material , and when a weight function for the detection is considered, the δ function δ (ξ) given in Fig. 19 is also an example thereof. This, however, cannot bring the size of visible area of the two optical systems the same because the cameras are installed at different receiving angles, respectively. If the cameras are installed for measuring a diffuse reflection light, the change in the weight function is not easy because the camera positions have to be changed.
For the former issue, measurement on the same optical axis is required. And, it is preferred that both components of the specular reflection and of the specular-diffuse reflection are grasped by the measurement in the direction of regular reflection on the steel sheet, not grasping the diffuse reflection light. For the latter issue, it is preferred that the weight function can be set with some degree of freedom against the changes in the camera position.
According to the object, the present invention adopts a linear light source having a diffusional characteristic, not a parallel light source such as that of laser light. Furthermore, the specular reflection component and the specular-diffuse reflection component are separated and identified from the regular reflection direction to the steel sheet using polarized light.
To explain the action and the effect of the linear diffusional light source, a linear diffusional light source 114 is placed in parallel with a steel sheet 104, as shown in Fig. 20, and the reflection characteristic is investigated by observing a point which is in a plane vertical to the light source and which is on the steel sheet 104 from the direction that the incident angle coincides with the outgoing angle, (hereinafter referred to as the "regular reflection direction to the steel sheet").
As shown in Fig. 20(a), when the light is emitted from center part of the linear light source 114, the light entered the tempered part is reflected in a specular mode, all of which is caught in the regular reflection direction to the steel sheet. On the other hand, the light entered the non-tempered part is reflected in a specular diffusional mode, of which only the light reflected from micro-area elements that face the same direction with that of the normal to steel sheet can be detected. Since the number of those micro-area elements is very few in probability, the reflected light that is detected in the regular reflection direction to the steel sheet is occupied mainly by the specular reflection from the tempered part.
To the contrary, when the light is emitted from a part other than the center part of the linear light source, as shown in Fig. 20 (b), the light entered the tempered part is reflected to a direction other than the regular reflection direction to the steel sheet by specular reflection, thus the light cannot be detected in the regular reflection direction to the steel sheet. On the other hand, the light entered the non-tempered part is reflected in specular-diffuse reflection mode, of which the light reflected in regular reflection direction to the steel sheet can be detected. Consequently, all the reflected light that can be detected in regular reflection direction to the steel sheet is the light of specular- diffuse reflection on the non-tempered part.
Both of the above-described cases lead a conclusion that, for the light emitted from the whole area of a linear light source, the detective light under observation from regular reflection direction to the steel sheet is the sum of the specular reflection light on the tempered part and the specular-diffuse reflection light on the non-tempered part.
The following is the description about the variations in polarized light characteristics on observation on an inspection plane from the regular reflection direction using that type of linear light source.
Generally, for the reflection on a specular metal surface, particularly to a light having the direction of electric field in parallel with the incident plane, (or p-polarized light), or to a light normal to the incident plane, (or s-polarized light), the polarized light characteristics are maintained after the reflection, thus the p-polarized light outgoes as the p-polarized mode, and the s-polarized light outgoes as the s-polarized mode. An arbitrary linear polarized light that has p-polarized component and s-polarized component at a time outgoes as an elliptically polarized light responding to the reflectance ratio and the phase difference of p- and s-polarized lights.
The following discusses the case that a light is emitted from a linear diffusional light source onto an alloyed galvanized steel sheet. As shown in Fig. 21(a), the light emitted from center part of the linear light source 114 is specularly reflected at the tempered part of the steel sheet 104 and is observed in the regular reflection direction to the steel sheet. In this case, the ordinary reflection on a specular metal surface is established, thus the p-polarized light outgoes as the p-polarized mode.
On the other hand, the light emitted from a part other than the center part of the linear light source is specularly reflected on micro-area elements that are inclined on the crystal surface on the non-tempered part, as shown in Fig. 21 (b), thus, some of the reflected light can be observed in the regular reflection direction to the steel sheet. In this case, even when a p-polarized light parallel with the incident plane of the steel sheet is entered, the light becomes a linear polarized light having both p- and s-polarized light components because the incident light is not in parallel with the incident plane for the micro-area elements which are inclined from which the light is actually reflected. As a result, the incident light outgoes from micro-area elements as an elliptical polarized light. The same result appears when an s-polarized light is entered instead of p-polarized light.
As for a linear polarized light with an arbitrary polarization angle, having both p- and s-polarized light components, the same reason as above-described can be applied, or, since the polarization angle becomes inclination from the incident plane, the shape of elliptical polarized light that is emitted in the regular reflection direction to the steel sheet differs from that of the light that entered from the center part of the linear light source and is reflected from the tempered part.
For the case of emitting linear polarized light having both p- and s-polarized light components, more detail explanation is given below.
As shown in Fig. 22, a light 108 coming from the linear diffusional light source 114 is converted to a linear polarized light by a sheet polarizer 115 having an azimuth α, which is then entered the steel sheet 104 positioned in horizontal direction. The regular reflection light is received by a light detector 116.
As described before, for the light 108 emitted from a point C on the light source, both the component specularly reflected from the tempered part and the component reflected in specular-diffuse reflection mode from micro-area elements that, by chance, the normal thereof directs to the vertical direction in the non-tempered part contribute to the light reflected from the point ○ (and from a region 113 peripheral to the point ○) on the steel sheet to the direction of the light detector 116.
To the contrary, as for the light 108 emitted from the point A which is offset by an angle φ viewed from the point 0, the specularly reflected light component is reflected in a direction different from that of the light detector 116, thus only the component of specular-diffuse reflection on micro-area elements with a normal angle ξ (the angle of normal to the vertical direction is ξ) contributes. The relation between φ and ξ is given by the equation below under a simple geometrical consideration. $\cos ξ = 2 \cos \cdot θ \cos^{2} (φ / 2) / [\sin^{2} φ + 4 \cdot \{\cos^{2} {θ \cos}^{2} \cdot (φ / 2) + \sin^{2} θ \cdot \sin^{4} (φ / 2)\}]^{1 / 2}$
Where, θ designates the incident angle to the steel sheet.
The state of polarized light of the light reflected in that manner is considered in the following. Referring to Fig. 22, the light 108 which is emitted from the point C passes through the sheet polarizer 115 having an azimuth α, then is reflected at the point ○ on the steel sheet. The polarized light state at that moment is expressed by Jones matrix which is generally used in the polarization optics. $E c = T \cdot E in$

where, E in designates the linear polarized light vector (column vector) at an azimuth a , and T designates the reflection characteristic matrix of the steel sheet. The component for each of them is written as follows. $\begin{array}{l} E in = Ep \cdot (\cos α, sinα) \\ T = r_{p} (T_{mn}); T_{11} = tanΨ \cdot \exp (jΔ), T_{22} = 1, T_{12} = T_{21} = 0 \end{array}$

where, '( ) designates the column vector, tanyψ designates the amplitude reflectance ratio of p- and s-polarized lights, Δ designates the phase difference occurred from the reflectance of p- and s-polarized lights, and r_s designates the s-polarized light reflectance. The matrix expression of them becomes the formula 1. $\begin{array}{l} E in = Ep (\begin{matrix} cosα \\ sinα \end{matrix}) \\ T = r_{z} (\begin{array}{l} tanψ \cdot \exp (jΔ) & 1 \\ 1 & 0 \end{array}) \end{array}$
In a similar manner, referring to Fig. 23, the polarized state of light 108 emitted from the point A and reflected on micro-area elements having normal angle ξ to the direction of light detector 116 is expressed by eq. (3) under an assumption that the incident plane crosses orthogonally with the sheet polarizer 115 and an analyzer 117. $E A = R (ξ) \cdot T \cdot R (- ξ) \cdot E in$

where, R designates the two-dimensional rotary matrix, and the component R_mn is expressed by: $R_{11} = R_{22} = \cos ξ, R_{11} = R_{22} = - \sin ξ$
The matrix expression of R(ξ) becomes the formula 2. $R (ξ) = (\begin{matrix} \cos ξ & - \sin ξ \\ \sin ξ & \cos ξ \end{matrix})$
Eq.(2) is a particular case of eq. (3) substituting ξ=0. Thus, both the specular reflection component and the specular-diffuse reflection component can be integrally treated by eq.(3).
When eq.(3) is calculated to draw a figure of elliptical polarized light state for the light reflected from micro-area elements having a normal angle ξ. Fig. 24 is obtained. The azimuth a of the incident polarized light was assumed to 45 degrees, the incident angle θ was assumed to 60 degrees, and the reflection characteristics of steel sheet were assumed as ψ=28° and Δ=120° The figure suggests that the ellipse inclines with variations in ξ value against the ellipse at ξ = 0 or against the case of specular reflection. Consequently, for example, by inserting an analyzer before the light detector to set the analyzing angle, selection becomes possible to determine the main reflected light coming from particular micro-area elements with a particular normal angle.
To quantify the above-described procedure, the state of polarized light E_p, which is obtained by inserting an analyzer having an analyzing angle β into a reflected light in a polarized state, is expressed by eq.(3). $\begin{array}{l} E D = R (β) \cdot A \cdot R (- β) \cdot E A \\ = R (β) \cdot A \cdot R (- β) \cdot R (ξ) \cdot T \cdot R (- ξ) \cdot E in \end{array}$

where, A = (A_mn) designates the matrix expressing the analyzer, and A₁₁ = 1, while other components are 0. The matrix expression of A becomes the formula 3. $A = (\begin{matrix} 1 & 0 \\ 0 & 0 \end{matrix})$
When the light intensity L of the reflected light on the micro-area elements having a normal angle ξ, detected by the light detector 116 (Fig. 23) is calculated by eq. (4), the light intensity L is expressed by eq. (5) with an assumption of the area percentage of the micro-area element of S(ξ). $\begin{array}{l} L = S (ξ) \cdot | E D |^{2} = {r_{B}}^{2} \cdot {Ep}^{2} \cdot S (ξ) \cdot I (ξ β) \\ I (ξ β) = \tan^{2} ψ \cdot \cos^{2} (ξ - α) \cdot \cos^{2} (ξ - β) + 2 \cdot tanψ \cdot cosΔ \cdot \cos (ξ - α) \cdot \sin (ξ - α) \cdot \cos (ξ - β) \cdot \sin (ξ - β) \cdot \sin^{2} (ξ - α) \cdot \sin^{2} (β - ξ) \end{array}$

where, I (ξ, β) is, as described before, the weight function that determines the degree of identification of reflected light on the micro-area elements having a normal angle ξ, which weight function depends on the polarization characteristics of optical system and of inspection body. The product of the weight function and the reflectance of steel sheet, r_s ², the incident light quantity Ep² , and the area rate S(ξ) is the light intensity that can be detected. In the case of a surface-treated steel sheet, or a homogeneous material on the surface of steel sheet, the value of r_s ² should be constant. In addition, the value of Ep² may also be constant if the incident light quantity is uniform at all positions of light source. Accordingly, to determine the light intensity that is detected by the light detector, only variables to be considered are the area percentage S(ξ) of micro-area elements having a normal angle ξ and the identification characteristic I(ξ,β).
Regarding the identification characteristic I(ξ,β), when an analyzing angle β_o that makes the contribution of the micro-area elements having a normal angle ξ_o maximum is selected, the candidates can be given by solving eq.(6) in terms of β. $[\partial I (ξ β) / \partial ξ] ξ = ξ_{0} = 0$
The general arithmetic expression of eq.(4) is given by the formula 4. ${(\frac{\partial I (ξ β)}{\partial ξ})}_{ξ = ξ_{0}} = 0$
When the analyzing angle that gives ξ = 0, or that gives maximum contribution of the specular reflection component is determined by eq. (6), the value of β becomes around -45 degrees. Also in this case, the reflection characteristics of the steel sheet adopted ψ = 28° and Δ = 120° , and the azimuth of polarized light α was 45°. Fig. 25 shows the relation between the normal angle ξ to the vertical direction of micro-area element and the identification characteristic, or the weight function I(ξ,-45), in the case that the analyzing angle β is -45 degrees. For convenience of visibility, the maximum value is standardized to 1.
Fig. 25 shows that the ξ= 0, or the specular reflection component, is the governing angle (easy for identification), and that the specular-diffuse reflection light on micro-area elements around normal angles of ξ = ± 35 degrees is most difficult to be identified. Inversely, an analyzing angle β that the reflection light at ξ = ± 35 degrees is identified best is determined from egs. (5) and (6), and the value of β becomes around 45 degrees. Fig. 26 shows the relation between the normal angle ξ against the analyzing angle β = 45 degrees and the identification characteristic I (ξ, 45). The curve of β = 45 degrees is not symmetrical in right and left sides. This is a result of that, in view of incident light plane (flat plane formed by the incident light and the reflected light relating to the micro-area element), a positive value of ξ gives apparently less azimuth α of the incident polarized light, (or becomes close to p-polarized light), and that the reflectance of p-polarized light on the steel sheet is less than the reflectance of s-polarized light. Fig. 26 also shows the case of β = 90° which gives an intermediate characteristic between β = -45° and 45° .
As given in eq. (5), the reflected light intensity L on a micro-area element having a normal angle ξ is given by a product of the identification characteristics (weight function) I(ξ, β) and the area percentage S(ξ). Accordingly, the intensity of the light received by the light detector 116 is the integrated value of S(ξ)I(ξ,β) in terms of ξ. For example, when a reflected light on a steel sheet having the reflection characteristics shown in Fig. 27 is received through an analyzer having an analyzing angle of β = -45 degrees, the quantity of received light is the integration of the area percentage S(ξ) shown in Fig. 27 with a weight of identification characteristics I(ξ,β) shown in Fig. 25.
If a pattern-like scab having characteristics shown in Fig. 16 exists, the area percentage S(ξ) becomes respective Figs. 17(a), (b), and (c).
For the case that only the specular reflection component is different as shown in Fig. 16(b) and Fig. 17(b), the light intensity on receiving that type of flaw through an analyzer having an analyzing angle β = -45 degrees corresponds to the result of integration of Fig. 17(b) multiplied by a weight function I(ξ,β) expressed by Fig. 25. Therefore, the difference in the reflected light quantity between the mother material and the scabbed portion can be detected. Regarding the analyzing angle β = 45 degrees, there is no difference in the specular-diffuse reflection component, as shown in Fig. 17(b), and the difference appears only at nearby ξ = 0° . Therefore, considering that the weight function I(ξ,β) at β = 45° given in Fig. 26 is a low value at around β = 0°, the product becomes a low value over the whole range of ξ, and the difference is cancelled by integration. As a result, no difference between the mother material and the scabbed part can be detected.
In the case that the difference appears only the specular-diffuse reflection component, as shown in Fig. 16(c) and Fig. 17(c), the detection cannot be attained by passing through an analyzer of -45 degrees. In that case, the detection can be done by passing through an analyzer of 45 degrees that provides high value of weight function I(ξ,β) distant from β = 0°.
The normal angle ξ giving no difference in the specular-diffuse reflection component between the mother material and the scabbed portion is around ξ = ± 20 degrees in Fig. 17 (c). If, however, there is a flaw that gives normal angle ξ nearby ± 30 degrees, the flaw cannot be detected even through an analyzer of 45 degrees. In that case, a separate analyzing angle (for example, β = 90°) providing different identification characteristic is prepared, and the light is received by the third light detector.
Generally, in most cases, the reflection characteristic of the mother material and scabbed portion on the surface of steel sheet falls in either one of Figs. 10 (a), (b), and (c). Accordingly, detection can be done in most cases by applying either two of the optical conditions (in this example, the analyzing angle). In a special case as described above, however, to prevent overlooking, it is preferable to use three analyzers each having different analyzing angle from each other and to receive the light by identifying the reflected light on micro-area elements having respective three normal angles.
When there is a difference in both the specular reflection component and the specular-diffuse reflection component, as in the case of Fig. 16(a) and Fig. 17(a), basically the difference between the mother material and the scabbed portion can be detected only from the reflected light passed through a single analyzer.
An incident sheet polarizer is located covering the whole area of a linear diffusional light source, and the azimuth of the polarized light includes both the p-polarized light and the s-polarized light. Furthermore, there adopt a camera to take image via a polarizer having a polarizing angle further penetrating the specular reflection component in the regular reflection light, and a camera to take image via a polarizer having a polarizing angle further penetrating the specular-diffuse reflection component.
This type of optical system conducts observation along a common light axis in the regular reflection direction, so that two kinds of signal are available corresponding to respective specular reflection and specular-diffuse reflection without being influenced by the variations of distance of steel sheet and by the variations of speed. Thus, a surface flaw inspection device that can detect pattern-like scab having no significant surface irregularity is realized. The detection range of angles for specular-diffuse reflection component is readily changed by determining the analyzing angle.
Furthermore, by determining the intensity or rate of the specular reflection and the specular-diffuse reflection, changes in surface property that affect the specular reflection or the specular-diffuse reflection, other than the above-described pattern-like scab, can be detected. For example, for the surface finish of metal strip, such as dull finish and hairline finish, can be detected, in theory, if only there is a variation in distribution of micro-reflection-face, and the application to inspect that type of surface property is expected.
The detection and the judgment of surface flaws may naturally apply known method and means in parallel. The detail of the parallel application of known method and means is described later.
In this manner, the position of the inspection plane that is judged to have a surface flaw is tracked by a tracking means. The tracking can be conducted by calculating the time that the position of surface flaw reaches the marking means, on the basis of the transfer speed of the metal strip. The marking means applies marking on the surface of the metal strip based on the marking command generated :from the tracking means.
Marking can be done by various methods depending on object and use. Any kind of marking method may be applied if only the marking is readily detected in succeeding stage. For example, printing by ink or paint, stamping using a stamper, drilling using a drilling machine, change of surface roughness using grinder or the like can be applied. For the case of ferromagnetic metal strip, a magnetic marking or the like can be applied.
The position of marking may be matched with the position of surface flaw, or may be matched thereto only in the longitudinal direction, not in the width direction. For example, if automatic feeding to a press-line as a material is adopted, the detection of marking may, in some cases, become easy by setting the marking position to a fixed position rather in width direction.
Herein is disclosed a method for manufacturing metal strip with marking, according to claim 3.
According to the method, a marking is applied to the surface of metal strip at the place where a surface flaw is judged as existing by the above-described surface flaw judging method. Since the marking to indicate the presence of surface flaw is applied, succeeding stage or user can remove the portion of the surface flaw, thus preventing the defect portion from entering the products. With the manufacturing method, the work of coil dividing to remove the surface flaw portion is significantly simplified or is eliminated, so that the production efficiency improves.
In an embodiment, the following method steps are carried out: winding the marked metal strip to prepare a coil; rewinding the coil to detect marking: avoiding or removing a specific range of the metal strip based on the information given by the marking; and applying specified working to a residual portion of the metal strip after avoiding or removing the specified range.
According to this embodiment, marking is applied onto the surface of metal strip, and the metal strip is wound to form a coil. The coil is transported to a plant or the like, where the forming-work is applied to produce steel sheet. On applying the forming-work, the coil is unwound in advance to detect a marking by visual inspection or using a simple detector. When the marking is detected, the defect portion including the flaw on the metal strip is avoided or removed based on the information.
For example, when marking is applied matching the position of flaw, the range of the defect portion is the portion applied by marking. When the marking has information of kind, degree, or the like of the flaw, the determination is given on the basis of the kind and degree of flaw which becomes a defect during the forming-work. The phrase "the defect portion including the flaw on the metal strip is avoided or removed based on the information" means that the defect portion of the metal strip is cut to remove, or the feed of the metal strip to the working stage is adjusted to pass the defect portion of the metal strip, thus controlling the feed of the metal strip to the working stage not to work the defect portion.
Herein is disclosed a metal strip with marking having, on a portion that shows an abnormality compared with a portion of normal combination of surface reflected light components under two or more optical conditions different from each other, the marking indicating information relating to a flaw on the surface thereof.
The metal strip is applied with marking at a place where the above-described surface optical analysis judged as not normal, or the position of surface flaw. Accordingly, as described above, succeeding stage or user of the metal strip can remove and prevent the portion of the abnormal part from entering the products.
Furthermore, marking may be applied to the surface relating to the surface inspection result or the information of various surface properties, based on the ordinary surface flaw inspection in terms of flaw size and shape, or reflectivity of emitted light, or the like. The "abnormal part" referred in the third aspect of the Best Mode means the part that, when reflected lights are separated under two or more of optical conditions, as described above, the intensity or the ratio of the reflection component differs from that of the normal part.
Another metal strip is disclosed, with marking having, on a portion that gives an abnormal quantity of light for one or both components of a specular reflection component on surface and a specular-diffuse reflection component on plurality of micro-area reflection surfaces, the marking indicating information relating thereto.
The other metal strip has a marking at a position where the state of specular reflection or of specular-diffuse reflection on the surface differs from that of normal portion. The term "specular-diffuse reflection" means the plane on which plurality of micro-area specular reflection planes on which the normal faces to a specified direction are distributed. Similar with the above-described aspects, the treatment of abnormal part becomes easy with the use of the metal strip.
The metal strip may have, about the information relating to the metal strip surface containing a portion that gives an abnormal quantity of light for one or both components of a specular reflection component on surface and a specular-diffuse reflection component on plurality of micro-area reflection surfaces, marking applied on the surface to indicate the information relating thereto.
Marking may be applied to the surface relating to the surface inspection result or the information of various surface properties, based on the ordinary surface flaw inspection in terms of flaw size and shape, or reflectivity of emitted light, or the like. The "abnormal part" referred in the fourth aspect of the Best Mode means the part that, as described above, the state of specular reflection or specular-diffuse reflection on the surface differs from that of normal part, and, when a reflected light is separated under two or more of polarization conditions, the intensity or the ratio of the reflection component differs from the normal part.
With the above-described aspects of the metal strip, the marking indicating the information about the abnormal parts of various surface flaws including abnormality in specular-diffuse reflection or about the abnormal parts of surface property is applied on the surface of metal strip. Accordingly, succeeding stage or user can notice the kind and degree of the surface flaw, thus being capable of responding to various uses and objects.
Furthermore, by applying marking on the surface of metal strip, the metal strip can be wound without cutting-off the surface flaw portion and other defective portions, which prevents from increasing the number of coils by strip cutting Since the number of coils is not increased, the coil handling does not increase the winding work. In addition, during transfer, rewinding, and working on the coils, the handling work is reduced because the number of coils is not increased.
A flaw inspection device for a metal strip is disclosed, according to claim 1.
The flaw inspection device having the light-receiving part and the signal processing section, is combined with an ordinary surface inspection means that inspects abnormality of the surface property such as flaw and stain by detecting size and shape of flaw and stain, or reflectance of the emitted light, or the like, thus classifying the kind and degree of abnormal portions such as surface flaw. By the procedure, total judgment is given on various kinds of abnormalities in surface properties such as abnormal specular-diffuse reflection, thus the marking of the information about these abnormalities is available.
Fig. 1 shows a block diagram of an example of carrying out the present invention. A surface flaw detection device 141 identifies a light reflected from the metal strip 104 under two or more optical conditions different from each other. A signal processing section 130 judges the presence/absence of surface flaw on the inspection plane based on the combination of these reflection components.
A tracking means 143 calculates the time that the position of surface flaw arrives at a marking means. That is, a sheet length calculation means 147 coverts the position of the surface flaw into the sheet length on the basis of the rotational speed determined by a rotameter 146 attached to a transfer roll 145, and converts the covered sheet length into the time necessary to arrive at a marking means 144. When thus determined time comes, the tracking means 143 generates a command signal for marking to the marking means 144. On receiving the command, the marking means 144 applies marking on the surface of the metal strip to indicate the position by printing, drilling, or the like.
Fig. 2 shows an example of the metal strip with marking. According to the example, the position of a marking 149 matches the position of surface flaw 111 in longitudinal direction, and maintains a fixed position from an edge in the width direction. Accordingly, for applying in a press line, the marking 149 can be detected at a fixed position from an edge independent of the position of the surface flaw 111, and it is possible to give treatment such as rejection of a certain portion including the surface flaw 111, thus preventing the production of defective products.
Fig. 3 and Fig. 4 show an example of the surface flaw detection device 141. As a linear diffusion light source 122, a transparent light-conductive rod applied with a diffuse reflection paint on a part thereof is used. A light emitted from a metal-halide light source is entered to both ends of the transparent light-conductive rod. The light coming out from the light-conductive rod of a light source 122 in diffusional mode passes through a cylindrical lens 125 and a sheet polarizer 126 with 45° polarization, then is conversed in a line with 60° of incident angle to enter over the whole width of a steel sheet 121. A reflected light 127 is further reflected by a mirror 128 located in regular reflection direction to the steel sheet, and enters camera units 129a through d, structuring the light-receiving part.
These camera units 129a through d are arranged in the sheet width direction, as shown in Fig. 5. With that positioned mirror 128, the facility can be designed in compact size. When the mirror 128 is positioned at an adequately distant from the steel sheet 121, the mirror 128 gives a region that comes outside of the view-field of all cameras, as shown in Fig. 5, thus the mirror can be structured with divided segments. The divided mirror construction decreases the fabrication cost.
Each of the camera units 129a through d in the light-receiving part comprises three linear-array cameras 132a through c, having respective analyzers 133a through c with respective analyzing angles of -45°, 45°, and 90° in front of each lens, while the light axes are in parallel to each other. The offset of the view-field of these three cameras is compensated by a signal processing section 130. With the light axes kept in parallel to each other, respective individual pixels of the three cameras 132a through c agree one-to-one to each other within the same view-field. Compared with the division of a single reflected light using a beam splitter, the method avoids loss of light quantity, and efficient measurement is available.
The light-receiving range A of individual light-receiving cameras 132a through c in each of camera units 129a through 129d overlaps in a part with the light-receiving range A of the corresponding light-receiving cameras 132a through c in each of other adjacent camera units 129a through d, as shown in Fig. 5. In other words, the light reflected from arbitrary position in the width direction on the steel sheet 121 is received by at least one of the three kinds of light-receiving cameras 132a through c in each of the camera units 129a through d.
Instead of the linear array camera, the light-receiving part may use a two-dimensional CCD camera. In addition, the light-emitting part may use a fluorescent lamp as the linear diffusional light source 122. Furthermore, a fiber light source may be applied by arranging the light-emitting end of a bundle of fibers in a line. That is, since the light emitted from each fiber has sufficiently broad angle responding to the fiber N/A, the fiber light source arranged with the fibers substantially functions as a diffusional light source.
The detail of the arrangement of plurality of cameras is described referring to Fig. 5. The plurality of camera units 129a through d are arranged at a fixed spacing therebetween. Each of the camera units 129a through d comprises three cameras 132a through c which receives light under different conditions (polarization of -45° , 45° , and 90° , respectively). These cameras are arranged in parallel to each other at a fixed spacing therebetween. Accordingly, the view-field of each camera offsets by the amount of camera distance.
The sequent order of camera arrangement in every camera unit is the same thereeach. For example, 45° , 90° , and -45° from left to right viewed from front side thereof. The measuring range (effective range), for example, is defined as the range that is observed under three kinds of optical conditions. And, a range where observation can be available only under one condition or only under two conditions, (range on both end portions), is concluded as ineffective, and not to be used. The camera spacing and the unit spacing are determined as a value that allows the maximum width of steel sheet to enter the measurement range (effective range).
The three cameras in each unit are not adjusted to provide the same view-field. After each camera determined the flaw candidate region, each camera is adjusted in terms of each flaw candidate region. As described above, since the view-field of each camera is offset from each other, in some cases not all of these three cameras can have a view-field for a certain flaw candidate region, (or three optical conditions cannot be satisfied). In these cases, the three optical conditions are satisfied using the results of the cameras of adjacent unit. The concept is applicable not only for receiving light of three polarized lights, but also for observing under arbitrary two or more conditions by dividing the total width of inspection body into plurality of view-fields.
Hereinafter the plurality of light-receiving part and the signal processing section are referred to as the flaw inspection means. Then, the surface flaw marking device shown in Fig. 1 is redrawn to Fig. 7. The flaw inspection means 140 has the light-receiving parts 132a through c, (corresponding to the cameras in Fig. 5 and Fig. 6), and the signal processing section 130. The signal processing section 130 conducts signal processing to detect the above-described diffusion specular reflection component based on the intensity of the reflected light which is identified under different optical conditions, thus giving judgment of presence/absence of abnormal part. After that, similar with Fig. 1, the position of surface flaw is calculated using the tracking means 143 and the sheet length calculation means 147, and applies marking to the position of abnormal part using the marking means 144.
As for the signal processing section, Fig. 8 shows an example of block diagram. The light intensity signals a through c coming from respective light-receiving cameras 132a through c enter respective average value decimation parts 134a through c, thus calculating the average value. After that, based on the pulse signals entered along with the movement by a certain distance in the longitudinal direction of the inspection body, the signal for a single line in the width direction is generated. By the decimation treatment, the resolution in the longitudinal direction is maintained to a fixed value. In addition, if the frequency of calculation of average value is regulated so as the moving distance in the longitudinal direction of the inspection body to not come outside of the view-field of the light-receiving cameras 132a through c, overlooking can be avoided.
Then, pre-treatment sections 135a through c compensate the irregular luminance relating to signals. The irregular luminance referred herein includes that caused from optical system, that caused from reflectivity of inspection sheet. The pre-treatment sections 135a through c detect the edge position of the steel strip and apply treatment not to mis-recognize sudden changes in signal at edge part as a flaw.
The signals completed the pre-treatment enter binary calculation sections 136a through c, where flaw candidate points are identified by comparing with preliminarily set threshold value. The identified flaw candidate points enter characteristic quantity calculation sections 137a through c, where the signal processing for flaw judgment is conducted. In the case that the flaw candidate points are in a sequential mode, characteristic quantity calculation sections 137a through c calculate the position and the characteristic quantity of, for example, starting address and ending address, and further the concentration characteristic quantity such as peak value.
The calculated characteristic quantities enter a specular flaw judgment part 138a or a specular diffusional flaw judgment part 138b depending on the optical conditions (with an analyzing angle β) of the original signals a through c. The output of the characteristic quantity calculation section 137a comes from the optical condition of original signal a as -45° analyzing (β = -45°). In this case, the characteristic quantity enters the specular flaw judgment part 138a to detect the difference in reflected light quantity between the mother material and the scabbed portion based on the specular reflection component, as described above.
On the other hand, the output of the characteristic quantity calculation sections 137b and c come from the optical conditions of original signals b and c as 45° and 90° analyzing angles (β = 45° and 90°); respectively, giving difference only on the specular-diffuse reflection component. Thus, the characteristic quantity enters the specular diffusional flaw judgment part 138b to give flaw judgment on the specular-diffuse reflection component.
Finally, a flaw total judgment section 139 gives judgment on the kind and the degree of flaw on the inspection plane of the metal strip based on the output of the specular flaw judgment section 138a and the specular diffusional flaw judgment section 138b. At that moment, considering the overlap of view-field between cameras 132a through d and between camera units 129a through d, (Fig. 6), it is preferable that the result of flaw judgment based on the signals coming from cameras of adjacent camera unit is used, at need.
Fig. 9 shows an example of combination of the surface flaw inspection means that gives flaw judgment by detecting abnormality in the specular-diffuse reflection component and a surface flaw inspection means applying other method. The surface flaw inspection means 140a is the same with that shown in Fig. 7. That is, plurality of light-receiving parts 132a through c identify the reflected light under different optical conditions, and the signal processing section 130 detects the abnormality in the specular-diffuse reflection component to give flaw judgment.
The surface flaw inspection means of other method, 140b, may apply ordinary surface flaw inspection means such as a device with the method to give judgment by detecting surface flaw based on the size and shape of the flaw, or a device with the method to detect surface contamination and adhesion based on the reflectance of emitted light, or other variables. The surface inspection means 140b classifies the ordinary surface flaw and abnormality in surface property in terms of the kind and the degree thereof. The marking information preparation means 142 conducts total classification and ranking on various kinds of surface flaw and abnormality in surface property, including abnormality in specular-diffuse reflection, thus preparing the information for marking.
After that, the tracking means 143 and the sheet length calculation means 147 calculate the position of the surface flaw, similar with the procedure in Fig. 1. The marking means 144 applies marking to the position of abnormality based on the marking information. At that moment, preferably the information relating to the kind and degree of the surface flaw is given. The information preferably gives a detectable form expressing marking pattern, shape, strip width, or the like. If bar codes or OCR (optical character reader) are applied, further detail information can be marked.
As described above, by applying marking on the surface of metal strip, increase in the number of coils is prevented, so that the work efficiency improves during the handling of coils, including transportation and recoiling. Furthermore, during the working of metal strip, the metal strip can be fed continuously without stopping at the flaw portion, so that an efficient work is expected.

Figs. 10 and 11 show the observed results on the alloyed galvanized steel sheet in accordance with the embodiment of Fig. 3. Fig. 10 corresponds to the above-described Fig. 17(b), and Fig. 11 corresponds to Fig. 17(c). The measured flaws are the one in which the area rate in the tempered part is larger in the scabbed part than in the mother material, and the diffusional property in the non-tempered part is the same therebetween, (Fig. 17 (b)), and the one in which the area rate in the tempered part gives no difference therebetween, and the diffusional property differs therebetween, (Fig. 17 (c)). As for the flaw of Fig. 11 type, generally there exist angles that cannot be detected in the diffuse reflection direction. The measurement of two kinds of that type flaws, each having different angles from each other, was conducted. For comparison, the figure also shows the result of non-polarized light observation, conforming to conventional technology, on entering light with 60° of incident angle and on measuring the light from regular reflection direction (60°) and from light-receiving angle (-40°) offsetting by 20° from the incident angle. The results are summarized in Table 1.

Table 1

Reflection characteristics at mother material and at scabbed part	Undetectable angle of receiving light	Light-receiving angle in accordance with conventional technology		Analyzing angle in embodiments
	Undetectable angle of receiving light	60	-40	-45	45	90
corresponding to Fig. 17(b)	-180 to 20°	○	×	○	×	Δ
corresponding to Fig. 17(c)	10 to 30, 55 to 65°	×	○	×	Δ	○
corresponding to Fig. 17(c)	-50 to -30, 55 to 65°	×	×	×	○	○

In Table 1, the symbol ○ designates detectable (large S/N value) and the symbol Δ designates undetectable (small S/N value).
Although the prior art adopts logic sum to receive light at two light-receiving angles and to remove noise, these flaws cannot be detected at two light-receiving angles at a time. Specifically, there are flaws that cannot be detected by either light-receiving angle.
To the contrary, according to the embodiment of the present invention, identification of the reflected light components corresponding to the three different light-receiving angles is done in the regular reflection direction by using analyzer. Accordingly, one of linear-array cameras can detect the flaws. Furthermore, it is easy to set optimum analyzing angle matching with the reflection characteristics of a flaw necessary to be detected.
Based on the finding that the reflection on the surface of steel sheet comprises the specular reflection component and the specular-diffuse reflection component, as described before, the present invention adopts a method to identify and grasp each component, which method comprises the steps of: using a linear diffusional light source; entering a polarized light having both p-polarized light and s-polarized light into the inspection plane; adequately setting the analyzing angle to the regular reflection direction to the steel sheet; thus identifying the component containing more of specular reflection component and the component containing more of specular-diffuse reflection component.
By the method, the unobservable flaw can be detected from the specular reflection component, and the pattern-like scabs having no significant surface irregularity, which cannot be detected in prior art, can be detected without fail. In addition, since both components can be grasped on the same light axis in the regular reflection direction to the steel sheet, the measurement free from the influence of variations in steel sheet distance and of variations of speed thereof has been realized. By setting the analyzing angle, selection of identification of the specular-diffuse reflection component at arbitrary angle has become available.
From the quality assurance point of view, that type of surface inspection device is absolutely required to not leave any non-detected flaw. The present invention actualizes, for the first time, the surface flaw marking device using a surface flaw inspection device that is applicable to wide fields including the surface-treated steel sheets without fail in detection of flaw, and the manufacturing of metal strips with marking. As a result, the surface flaw inspection that was relied on visual inspection of inspector is automated, and a simple means can notify the information to succeeding steps and to user. Thus, the use effect of the device and the method according to the present invention is significant.

Claims

A flaw inspection device (141), comprising:
a first light-receiving device that identifies a specular reflection component in a regular reflection light via a first polarizer, the first polarizer allowing the specular reflection component to penetrate therethrough more than a specular-diffuse reflection component, the regular reflection light having a regular reflection direction and coming from an inspection plane of a metal strip and having the specular reflection component and the specular-diffuse reflection component;

a second light-receiving device that identifies a specular-diffuse reflection component in the regular reflection light via a second polarizer, the second polarizer allowing the specular-diffuse reflection component to penetrate therethrough more than the specular reflection component; and

a signal processing device (130) that judges the presence/absence of a surface flaw on the inspection plane, based on the specular reflection component identified in the first light-receiving device and the specular-diffuse reflection component identified in the second light-receiving device,
characterized in that the first and second light-receiving devices are arranged to conduct observation along a common light axis in the regular reflection direction.
A defect marking device comprising:
flaw inspection means comprising a flaw inspection device (141) according to Claim 1; and

marking means (144) arranged to apply marking that indicates information relating to the flaw (111) on the surface of the metal strip (104).
A method for manufacturing a metal strip (104) with defect marking, comprising the steps of:
(a) a first identifying step of identifying a specular reflection component in a regular reflection light via a first polarizer, the first polarizer allowing the specular reflection component to penetrate therethrough more than a specular-diffuse reflection component, the regular reflection light having a regular reflection direction and coming from an inspection plane of a metal strip and having the specular reflection component and the specular-diffuse reflection component;

(b) a second identifying step of identifying a specular-diffuse reflection component in the regular reflection light via a second polarizer, the second polarizer allowing the specular-diffuse reflection component to penetrate therethrough more than the specular reflection component, wherein observation in the first and second identifying steps is conducted along a common light axis in the regular reflection direction;

(c) judging the presence/absence of a surface flaw on the inspection plane, based on the specular reflection component identified in the first light-receiving device and the specular-diffuse reflection component identified in the second light-receiving device; and

(d) marking information relating to the flaw on the surface of the metal strip based on the judgment result.
A method for working a metal strip according to Claim 3, further comprising the steps of:
(e) winding the marked metal strip to prepare a coil;

(f) rewinding the coil to detect the marking, and specifying a specific range of the metal strip based on the information given by the marking; and

(g) applying specified working to a residual portion of the metal strip after avoiding or removing the specified range.